Mechanism Design Theory and Sustainable Urban Form: A Proposed Priority for Collaborative Research Michael Mehaffy Abstract Recent news of the 2007 Nobel Prize in Economics has focused attention on a previously little‐known area of economic science known as mechanism design theory.1 It has been noted that this field holds promise for more sustainable forms of economic process, dealing with such urgent topics as climate change and resource depletion. Herein we consider the implications this field holds for the development of more efficient, higher‐ quality, more ecologically sustainable forms of urban settlement. We propose that to develop this potential, it will be important to pursue new collaborative forms of research between economists, urban planners and other disciplines. We note a particular opportunity for members of the urban reform movement known as The New Urbanism. Mechanism Design and Economic Game Theory It has been known for some time that markets work relatively efficiently to allocate resources based upon known information – Adam Smith’s so‐called “invisible hand” but do not work well to integrate so‐called “externalities”. These include environmental damages, services provided by ecological systems (purifying water for drinking, for example), long‐term social costs, and other factors. In addition it is now understood that human cognitive limits, asymmetric information and other forms of “bounded rationality” within market processes limit their efficiency and produce irrational outcomes.2 This “bounded rationality” has important implications for the ability of markets to deliver ecologically sustainable results. Mechanism design theory can correct these omissions by, in effect, “re‐calibrating” the economic process to adjust to the new factors. It does so through the design of “mechanisms” such as combinations of taxes, credits and other rule‐based processes that change the dynamic equilibrium of the market. Game Theory and Economic Behavior Mechanism design theory is in fact a form of game theory, that is, an understanding of the possible outcomes given a set of rules that, though they may be simple, may interact in a complex way (as, for example, the rules of Chess). Mechanism design theory asks, in effect, how can the rules of the game be changed so that the outcome is more like what we desire? We may want a game that does not continue on so long without a clear winner, or one that does not result in frequent stalemates, for example. Game theory applies not only to ordinary games, but to any situation in which the outcome depends upon the actions of independent agents, following a set of shared rules, such as traders in a market. The theory describes the tendency of certain rules to produce certain classes of outcomes. A famous example from game theory is the “prisoner’s dilemma.” In this example, two colleagues are charged with a crime, and questioned separately. The rules are set such that there are four possible outcomes. One, if Prisoner A confesses and also testifies against Prisoner B, and if Prisoner B does not confess and testify against Prisoner A, Prisoner B will get a harsh sentence and Prisoner A will get a light sentence. A second possible outcome is the converse: if Prisoner B testifies against Prisoner A but A stays silent, then Prisoner B will get the lighter sentence. 2 These effects are becoming well understood, and have been the subject of their own recent Nobel Prizewinning work. See for example the 2001 award to George A. Akerlof, A. Michael Spence and Joseph E. Stiglitz for information asymmetry, or the 2002 award to Daniel Kahneman and Vernon L. Smith for prospect theory. One famous study showed that drivers will readily spend an extra $1 on gasoline to drive to a “big‐ box” store that saves them perhaps $.79 on paper towels – an illusory but nonetheless behaviorally compelling bargain. In the third possible outcome, if each testifies against the other, each will get a somewhat harsh sentence, but not as harsh as the maximum. But in the fourth outcome, if neither testifies against the other, they will both go free. Here is the dilemma: not knowing what the other prisoner will do, what is the most logical choice for each prisoner? Clearly it would be best to say nothing and to have the other prisoner say nothing – then each will go free. But there is no way to be sure that the other prisoner will cooperate, in which case the most logical choice may be to hedge one’s bets and to testify against the other. Then at least one will get less than the harshest sentence, and perhaps get a much‐reduced sentence. Since both prisoners will tend to see this logic, each is very likely to testify against the other. Hence the tendency under such a set of rules – such a “game” ‐‐ is toward a condition in which both will tend to testify, in order to optimize their benefit. This overall condition is known as the game’s “Nash Equilibrium,” in honor of the theorist who identified it, John Nash (whose biography was featured in the book and movie “A Beautiful Mind”). Note, in this example, that changing the rules could change the Nash Equilibrium of the game. For example, giving the prisoners the same punishment no matter what they say would likely result in their not saying anything about their comrade, as they would not be rewarded for doing so. Mechanism Design Theory, then, is the study of how such sets of rules – such “mechanisms” ‐‐ can be designed such that the trend toward the optimum for all players in the outcome of the game – its Nash Equilibrium ‐ will change in a desired way. Mechanism Design Theory – An Elementary Example Mechanism Design Theory in the economic sphere seeks to change the rules of the economic game to shift the optimal outcome. Note that it does not in any way directly require that outcome. Rather, it shifts the tendency of the game to favor that outcome. One can readily see that mechanism design theory, even when employed by government, is not the same as direct government regulation – as, for example, a simple tax on cigarettes, or ban on recreational drugs, to discourage their use. It is a more subtle strategy for altering by design the optimal tendency for rule‐based outcomes of multiple economic interactions. A simple example from a domestic environment will serve to illustrate a simple but effective designed mechanism. Consider two young children who have been given a piece of cake and directed to share it. It is not unlikely that they will fight over the cake, or that one child (the bigger one perhaps) will get a larger piece of cake. To avoid this disparity, a parent may decide to try to direct the sharing of the cake, and may even intervene to cut the cake. This kind of action would be analogous to so‐ called “top‐down” government regulation. But consider an alternative strategy. The parent simply establishes a simple pair of rules: one child will cut the cake in two, and the other will have the first pick of the two slices. Both children know these rules, and can think through their consequences. The first child, told to cut the cake, may want to have a larger slice, but now knows that is impossible; the second child, having the first choice, will simply pick the larger slice. Therefore the only reasonable choice for the first child, in order to maximize the benefit, is to cut the slice as close to evenly as the child possibly can. The second child, facing two very nearly even slices, happily chooses one. As if by magic, the children have cut and shared equal pieces of cake peacefully. Thus the “rules of the game” have induced a fairer outcome than otherwise would occur, in a competition for limited and desirable resources. The normal tendency of the game to establish an optimum benefit for all players ‐‐ its “Nash Equilibrium” ‐‐ has been shifted by the new rules. The rules do not directly require a fair slicing; indeed, they are entirely silent on that issue. Rather, they create a set of conditions in which fair slicing is the most advantageous to all concerned. A mechanism has been designed that shifts the Nash Equilibrium of the cake‐sharing game, such that both children are rewarded for participating in an even slicing. The “Tragedy of the Commons” and Climate Change In the environmental realm, one can observe a phenomenon closely related to the Prisoner’s Dilemma. The so‐called “Tragedy of the Commons” refers to a hypothetical situation in which a group of villagers lives along a commons and shares a herd of cows. Each villager will be partially rewarded long‐term for milking the cows and ensuring that they are well cared for. But if any one villager slaughters a cow, that villager will be disproportionately rewarded for the temporary meat that the cow will provide – even though the benefit to the one villager is not as great as the sum benefits would have been to all villagers over time. Each villager knows that the other villagers know this as well, and hence each is eager not to be left without any cows to milk, or to eat. Hence the cows are quickly and almost inevitably slaughtered. The Nash Equilibrium of the game almost compels the villagers to do so. This phenomenon explains how, for example, Easter Islanders came to cut down the very last tree on their island centuries ago, surely knowing that these trees were essential for the canoes on which their fishing, and their livelihood, depended. How could such a seemingly illogical event have occurred? Yet clearly someone did cut down the last tree.